Speed, torque and thermal characteristics Most stepper motors are determined by their speed-torque characteristics, where torque is reduced by increased speed. Driving peristaltic pumps, especially with multiple channels in parallel requires significant torque. On the other hand, operating at flow rates at a wide range (e.g. 0.1 - 100 ml/min, which is suitable for dynamic cell culture experiments). Consequently, the motor used requires sufficient torque at the highest required speeds to allow appropriate perfusion control. However, an increased torque requires increased winding and thus typically increased motor sizes, which lead to more heat being generated by the motor. Consequently, it is necessary to select motors with the right balance of torque, speed and size to prevent overheating, especially within an incubator. For the purpose of this project, several motors were tested to determine sufficient torque in at speeds between 1 and 200 RPM (without microstepping) with the pump head and 8 cartridges with tubing installed. Nema 14, 23 and 34 step motors were tested in addition with a spare DC motor used in a Masterflex laboratory pump. Only the Masterflex spare DC and the Nema 34 step motors were sufficiently powerful to pump liquid at the highest rotation speeds.
Two set-ups were prepared, one using the Nema34 step motor and one using the Masterflex spare DC motor and run one after the other continuously inside a CO2 incubator for 1 week. Both motors performed well at cell culture conditions (37°C, 5% CO2 and 95% humidity). However, the larger, DC motor produced more excess heat, which caused condensation at the edges of the incubator door, which is less favourable. Thus the Nema 34 step motor was selected for further development.
Abrasion of the mechanical parts For maximum efficiency of torque transfer between motor and pump head, the motor and pump head axes need to be aligned and connected, which is typically done by 'coupler'. As perfect alignment is difficult to achieve, early tests have shown that direct 'metal-to-metal' coupling with slight 'off-center' movement of the components resulted of abrasion of the softer parts (e.g. aluminium coupler). However, this was solved by a rubber element (see figure below) which is placed between the coupler and the pump-head axis handle and absorbs movement deviations during rotation.
Abrasion of peristaltic tubing Perfusion experiments performed in the scope of this project have shown that peristaltic tubing was prone to abrasion during prolonged use which was especially prominent when using cell culture media (and less so when using PBS). Tubing spallation is a well known phenomenon and was described before, e.g. in clinical environments and in pharmaceutical research and development. Here, the tubing types summarized in the table below were used:
| n | tubing type | dimensions - IDxOD [mm] | producer/distributor | ref | specified sterilization |
|---|---|---|---|---|---|
| 1 | silicone | 1,5x3,5 | Carl Roth | 9557.1 | avtoklav |
| 2 | platinum cured silicone | 1,42x3,2 | Masterflex | 06421-34 | ? |
| 3 | BPT | 2,06x3,76 | Carl Roth | cyp7.1 | avtoklav |
| 4 | tygon | 1,587x3,175 | Masterflex | 06407-37 | ? |
| 5 | flexelene | 1,587x3,175 | Masterflex | 96181-00 | avtoklav |
| 6 | tygon | 1,52x3,22 | Carl Roth | cyt8.1 | avtoklav |
| 7 | silicone | 1x1,8 | Carl Roth | 9555.1 | avtoklav |
| 8 | tygon | 0,89x2,56 | Carl Roth | cyt7.1 | avtoklav |
All tubes were prone to spallation during use with cell culture medium (ADMEM) and left particulate matter in the culturing containers. A few examples are shown in the figure below, depicting bright-field photomicrographs of P6 well plates after ADMEM 5ml/min perfusion for 3hours, using 4 different tubing types.
As photosensitive resins suitable for DLP 3D printing were used to make some of the tools that come in touch with the cells and culture media (e.g. single well chips, hydrocyclone filters). To this end, several resins from 2 suppliers, namely W2P's Solflex resins and 3D Jake's Colormix basic with varying degrees of medical grade certification were tested for cytotoxicity. A cytotoxicity assessment was performed using the following materials:
| sample | supplier | certification |
|---|---|---|
| control (no resin) | N.A. | N.A. |
| Colormix basic | 3Djake | not specified |
| Solflex med clear | W2P | medical device class IIa |
| Solflex splint | W2P | medical device class IIa |
| Ortho clear | W2P | medical device class IIa |
| Splint comfort | W2P | medical device class IIa |
A Live/Dead assay and the overall metabolic activity of human umbilical vein endothelial cells (HUVEC) was evaluated using protocols as specified by reagent suppliers and previously described in this article. Here only a brief protocol with experimental specifics is described:
Resin sample preparation:
- Discs with 1mm thickness and 12.5mm diameter were LDP 3D printed, washed and cured (according to supplier specifications) for each resin type.
- The discs were sterilized by additional washing with 70% Ethanol and air dried under a laminar flow hood.
Cell-culture: Cell culture of HUVEC followed the protocol as described in the above referenced article. 3. The discs were transferred to P12 well plates (6 discs per resin type). 6 additional wells were kept empty for control. 4. 3 wells per material (resin and control) were covered with 1ml of ADMEM (+ 5% fetal bovine serum) and 20 000 HUVEC. 5. In parallel, the remaining wells were covered only with cell culture medium without the cells. 6. The samples were incubated at 37°C and 5% CO2 for 48h.
Alamar-blue assay: 7. The culture medium in all samples was replaced with fresh medium containing 16.7 μg/mL resazurin, followed by additional incubation for 2h at 37°C and 5% CO2. 8. After incubation the well plates were stirred to mix the medium containing reagent, followed by extraction 100μl aliquots to a fresh P96 well plate. 9. Fluorescence intensity of each sample was measured at an excitation wavelength of 530/25 nm and emission wavelength of 590/35 nm. 10. To evaluate the metabolic activity, the measured fluorescence from the mean value of controls (samples without cells, material specific) was subtracted from the mean value of respective test samples. The relative metabolic activity was determined by dividing the obtained values with the normalized value of the HUVEC cultivated in blank wells (no resin).
The results of the Alamar-blue assay are summarized in the figure below. All tested materials showed high metabolic activity of cells, with the highest values measured on the 'Colormix basic' resin, followed by 'Ortho clear'.
Live/Dead assay: The samples from point 8. were washed with fresh medium and covered with 0.5ml of ADMEM containing 5μg/ml Calcein (indicator of live cells - fluorescence in the green spectrum) and 20μg/ml Propidium Iodide (indicator of dead cells - fluorescence in the red spectrum) and incubated for 15min in darkness before fluorescence microcopy. Screen shots of 4 micrographs comparing the control sample with 3 resins are shown in the figure below. All samples show viable cells with almost none visible dead cells. Furthermore the optical properties of the resins can be estimated from the clarity of the images.
The results suggest good biocompatibility of the tested resins on HUVEC cells, however, it should be noted that these results are not representative for other cell type or set-up configurations. Thus, this type of analysis should be repeated for every experiment specifically.